Introducing W/Support, our state-of-the-art platform designed to transform customer service. At the heart of W/Support is SupportSage, an AI-driven support chatbot that seamlessly integrates to provide an interactive FAQ system and efficient chat interactions. SupportSage handles surface-level inquiries, allowing IT support technicians to concentrate on more complex issues, thus enhancing customer satisfaction and reducing agent workload.

• Introduction
• Project Goals
• Technical Overview
  • Architecture
  • Components
    • Frontend (Angular)
    • Proxy (Express.js)
    • Backend (Rust)
• Local Development
  • Prerequisites
  • Installation
  • Running the Application
  • Testing
• Contact

SupportSage is designed to provide efficient and accurate support to Hisense customers. By leveraging a large language model (LLM) and an interactive FAQ system, the chatbot can understand and respond to user queries, troubleshoot common issues, and guide users to relevant solutions. This will significantly improve the user experience, reduce response times, and free up valuable resources for more complex support issues.

• Enhanced Customer Support: Provide 24/7 availability and faster response times to customer inquiries.
• Reduced Support Costs: Lower the workload on IT support staff by automating responses to common questions.
• Improved User Experience: Deliver a personalized and intuitive support experience through natural language interactions.
• Scalability: Design a system that can easily adapt to increasing user demand and expanding knowledge bases.

SupportSage is built using a modern technology stack that combines the strengths of various programming languages and frameworks:

The system follows a microservice architecture, where each component has a distinct responsibility:

graph TD
  subgraph Frontend
    A["Angular Frontend"] -->|POST Request with Message| B["Express.js Proxy"]
  end

  subgraph Proxy
    B["Express.js Proxy"] -->|Forwards Request| C["Rust Backend: main"]
    D["Rust Backend: handle_chat_message"] -->|Response| B["Express.js Proxy"]
    B["Express.js Proxy"] -->|Response| A["Angular Frontend"]
  end

  subgraph Backend
    C["Rust Backend: main"] -->|Start Server, Setup CORS| D["Rust Backend: handle_chat_message"]
    D["Rust Backend: handle_chat_message"] --> E["Extract Message from Request"]
    E["Extract Message from Request"] --> F["Create Ollama Generation Request"]
    F["Create Ollama Generation Request"] --> G["Stream Response from Ollama"]
    G["Stream Response from Ollama"] --> H{Is Stream Complete?}
    H{Is Stream Complete?} -->|No| G["Stream Response from Ollama"]
    H{Is Stream Complete?} -->|Yes| I["Create JSON Response (ChatMessage)"]
    I["Create JSON Response (ChatMessage)"] --> D["Rust Backend: handle_chat_message"]
  end

The Angular frontend is responsible for:

• User Interface: Displaying the chat interface, user messages, and chatbot responses.
• Input Handling: Capturing user input and sending it to the backend for processing.
• Response Display: Rendering the AI-generated responses in the chat window.

The Express.js proxy serves as the intermediary between the frontend and backend. It handles:

• Request Routing: Directing incoming requests to the appropriate backend service.
• Load Balancing (Optional): Distributing requests across multiple backend instances for improved performance.
• Security: Implementing security measures like authentication, authorization, and rate limiting.

The Rust backend is the core of SupportSage, handling the AI interactions. It includes:

• Message Processing: Receiving user messages from the proxy, processing them with the Ollama LLM, and generating responses.
• Error Handling: Managing errors and exceptions that may occur during processing.
• CORS Handling: Enabling cross-origin resource sharing for secure communication with the frontend.

• POST /api/v1/chat: This endpoint handles incoming chat messages from the frontend and returns the AI generated responses.
  • URL: http://127.0.0.1:8080/api/v1/chat
  • Method: POST
  • Headers: Content-Type: application/json
  • Body:

    {
      "message": "Your message here"
    }

  • Response:

    {
      "message": "Your message here",
      "ai_response": "AI's response here"
    }

Ensure you have the following installed on your development machine:

• Node.js and npm/yarn: For running the Express.js proxy.
• Rust Toolchain: Rustup is recommended for installing and managing Rust.
• Ollama: Install and configure Ollama according to their official documentation.

1. Clone the frontend and backend repositories
2. Install dependencies

Start the Rust Backend:

cd support-sage-backend
cargo run

Start the Express.js Proxy:

cd support-sage-frontend
npm start
# or
yarn start

You can use tools like Postman to send requests to the /api/v1/chat endpoint with different messages and verify that the backend responds correctly. Refer to the Testing with Postman section in the Rust backend README for detailed instructions.

For any questions or inquiries, please contact us at our school e-mails!